From 998d15bca14b42efe5469532b9cf9ee2307fe175 Mon Sep 17 00:00:00 2001
From: lih <dev_lih@139.com>
Date: Thu, 20 Nov 2025 15:58:48 +0800
Subject: [PATCH] Fix: missing parameters in by_plaintext method for PDF naive
 mode

---
 rag/app/book.py         | 1 +
 rag/app/laws.py         | 1 +
 rag/app/manual.py       | 1 +
 rag/app/naive.py        | 8 +-------
 rag/app/one.py          | 1 +
 rag/app/presentation.py | 1 +
 6 files changed, 6 insertions(+), 7 deletions(-)

diff --git a/rag/app/book.py b/rag/app/book.py
index 5ea28d40d..5bdaec72d 100644
--- a/rag/app/book.py
+++ b/rag/app/book.py
@@ -113,6 +113,7 @@ def chunk(filename, binary=None, from_page=0, to_page=100000,
             lang = lang,
             callback = callback,
             pdf_cls = Pdf,
+            layout_recognizer = layout_recognizer,
             **kwargs
         )
 
diff --git a/rag/app/laws.py b/rag/app/laws.py
index dd97e4e3a..ba2592833 100644
--- a/rag/app/laws.py
+++ b/rag/app/laws.py
@@ -172,6 +172,7 @@ def chunk(filename, binary=None, from_page=0, to_page=100000,
             lang = lang,
             callback = callback,
             pdf_cls = Pdf,
+            layout_recognizer = layout_recognizer,
             **kwargs
         )
 
diff --git a/rag/app/manual.py b/rag/app/manual.py
index 5808e2498..4f9de40c7 100644
--- a/rag/app/manual.py
+++ b/rag/app/manual.py
@@ -213,6 +213,7 @@ def chunk(filename, binary=None, from_page=0, to_page=100000,
             lang = lang,
             callback = callback,
             pdf_cls = Pdf,
+            layout_recognizer = layout_recognizer,
             **kwargs
         )
 
diff --git a/rag/app/naive.py b/rag/app/naive.py
index 099e2f6fd..49dca17af 100644
--- a/rag/app/naive.py
+++ b/rag/app/naive.py
@@ -111,13 +111,7 @@ def by_tcadp(filename, binary=None, from_page=0, to_page=100000, lang="Chinese",
 
 
 def by_plaintext(filename, binary=None, from_page=0, to_page=100000, callback=None, **kwargs):
-    parser_config = kwargs.get("parser_config", {})
-    layout_recognizer = parser_config.get("layout_recognize", "DeepDOC")
-
-    if isinstance(layout_recognizer, bool):
-        layout_recognizer = "DeepDOC" if layout_recognizer else "Plain Text"
-
-    if layout_recognizer == "Plain Text":
+    if kwargs.get("layout_recognizer", "") == "Plain Text":
         pdf_parser = PlainParser()
     else:
         vision_model = LLMBundle(kwargs["tenant_id"], LLMType.IMAGE2TEXT, llm_name=kwargs.get("layout_recognizer", ""), lang=kwargs.get("lang", "Chinese"))
diff --git a/rag/app/one.py b/rag/app/one.py
index 5574aaa51..7cd1bb785 100644
--- a/rag/app/one.py
+++ b/rag/app/one.py
@@ -99,6 +99,7 @@ def chunk(filename, binary=None, from_page=0, to_page=100000,
             lang = lang,
             callback = callback,
             pdf_cls = Pdf,
+            layout_recognizer = layout_recognizer,
             **kwargs
         )
 
diff --git a/rag/app/presentation.py b/rag/app/presentation.py
index cd1d308ec..6a872528f 100644
--- a/rag/app/presentation.py
+++ b/rag/app/presentation.py
@@ -142,6 +142,7 @@ def chunk(filename, binary=None, from_page=0, to_page=100000,
             lang = lang,
             callback = callback,
             pdf_cls = Pdf,
+            layout_recognizer = layout_recognizer,
             **kwargs
         )